SK
Computer Science × Data Science — UVA

Sanjay
Karunamoorthy

Use the bot icon above to ask me anything

If the chatbot errors, email sanjaykarun@gmail.com to wake up AstraDB.

About

Who I Am

I'm a UVA Alumn with a B.S. in Computer Science with a Minor in Data Science. I'm passionate about solving hard problems and building things that matter from production infrastructure, ML systems, and everything in between.

Skills

Languages

JavaPythonCJavaScriptTypeScriptHTML + CSSSQL

Technologies

SpringDjangoReactOpenSearchNext.jsNode.jsPyTorchMongoDBExpressKafkaGitREST API
Experience

Where I've Worked

University of Virginia

Undergraduate AI Researcher

Aug 2025 – Present
  • Training Reinforcement Learning agents to collaborate with humans to improve task performance using Proximal Policy Optimization. Conducting experiments in multi-agent settings to analyze cooperation, incentives, and emergent behaviors.

GEICO

Software Development Intern

Jun 2025 – Aug 2025
  • Pioneered an MVP dashboard for a Project Development Life Cycle management system based on Shift Left methodology; built a REST API with Django, PostgreSQL, TypeScript, and React — scoped to benefit 3,000+ engineers.
  • Partnered with 15+ Senior Directors, VPs, PMs, and SMEs to design a workflow optimization system with JSON integration, Redis caching, GraphQL, and Azure DevOps/Slack automation.
  • Implemented bidirectional synchronization and timed reloads to prevent race conditions.
  • Engineered an email automation script by pre-processing 17,000+ repos to reach 2,000+ system owners; used Power BI, Excel, and pandas for data organization and delivery.

SS Technology Consultants

Software Engineer Intern

May 2024 – Jul 2024
  • Built an efficient data processing solution using Spring Boot and Spring Batch on a 3M Project, handling 2M+ search data points from Salesforce and SharePoint.
  • Optimized big data pipelines with Apache Airflow and Kafka, achieving a 6% improvement in semantic search processing efficiency.
  • Leveraged Amazon Bedrock + Titan Text Embeddings for NLP and vectorized search for RAG.
  • Implemented kNN vectors (k=5) with lambda functions on OpenSearch; prompt-engineered the Anthropic Claude API for context-relevant responses.

SS Technology Consultants

Full Stack Engineer Intern

Dec 2023 – May 2024
  • Led a team of interns to build a REST API with JIRA-managed agile sprints; built CI/CD pipelines using Docker, Kubernetes, Jenkins + Bitbucket, cutting deployment time by 30%.
  • Designed a secure auth system with React, Express, Node.js, and MongoDB using JWT and Bcrypt — eliminating all major authentication vulnerabilities.
  • Automated monitoring tasks in Linux/Unix with Bash scripting to reduce manual setup time.
Projects

Things I've Built

Fabric

  • Utilizing LangGraph, LLMs, A2A, MCP, RAG, and ADK to create a suite of AI Agents to optimize my daily tasks.

React · Electron · Python · FastAPI · Pydantic · ChromaDB

DaVinci Learning Graph

  • Innovating a novel way to understand, creating a neural map of everything you learn.

Chord Studio

  • AI-powered music and chord generator, from thought to music in seconds.

Web Audio API · Next.js · Typescript · Tailwind CSS · OpenAI API

Siblings or Dating

  • CNN (MobileNetV2) using transfer learning to classify Reddit images as siblings or dating. Achieved 80% validation accuracy.

TensorFlow · Keras · Python · CNN · Django

Akuna Capital Quant Challenge

  • Python market-making framework integrating nonlinear optimization, stochastic modeling, and RL for adaptive market intelligence.

Python · RL · Nonlinear Optimization

Poker Bot

  • Monte Carlo counterfactual regret minimization to approximate Nash Equilibrium for 2-player Kuhn's poker.

Python

TenantTalk

  • Web app for rating landlords and property management, built with a team of 6 using Scrum methodology.

Django · PostgreSQL · Heroku · AWS S3

Coccidiosis Classification

  • CNN to detect Coccidiosis in chicken images using 1,000+ labeled training images.

Python · TensorFlow · Azure

Film Finder

  • Recommendation system integrating the TMDB API across 5,000+ movies using scikit-learn.

Python · NumPy · scikit-learn · TMDB API

EzCite

  • Lead dev using React and Next.js. Integrated Perplexity LLM and Puppeteer for optimal citation generation.

React · Next.js · Node.js · JavaScript

NLP Application

  • End-to-end Named Entity Recognition app using Stanford CoreNLP to detect names, cities, states, and countries.

Java · Spring Boot

European Option Pricing

  • Monte Carlo simulation pricing model for European vanilla options using GBM and Box-Muller, matching Black-Scholes.

C++

EDO.ai

  • Researching and developing models to predict diabetes using extensive blood test data.

PyTorch